Last year I ran a few tests, such as
medium gun on Wolf and
missile launchers on Shrike. After those I wanted to do a “wolf pack” test, using different fleet compositions of small hunter-killer-like ships. But since 0.95 was coming I waited. I have now done such a test, and I want to share the results. Also, this is kind of a followup to the recent thread
Tempest Nerf Options.
The intent here is to have a look at attack performance of the various tested ships, using just one specific loadout. So the scope is fairly limited. Ability to defend and distract is not considered. In order to measure attack performance, I ran 24 repetitions of each matchups, and wrote down the battle outcome and the time to win.
Before moving to the interesting bits, it is important to keep in mind:
- this post is not meant to be a guide teaching how to play
- this is AI vs AI, player is not involved during battle
- this is about base attack performance, so no SO hullmod, no built-in hullmod, no officer, no skill
- Starsector 0.95-RC15 with a customised Fleet Tester mod: I removed the 5 capture objectives, and kept the 10000x10000 arena
- all ships had aggressive personality
- ship performance will change depending on its loadout, allied fleet composition, enemy fleet composition and enemy ship loadout
- obviously I did not simulate all possible battle situations
I ran 9 series of tests: 1.1, 1.2, 1.3, 2.1, 2.2, 3.1, 3.2, 4.1, 4.2. Details further below.
Above I wrote “using just one specific loadout”, but it is not exactly true. I started with vanilla variants for all tested ships (1.1), then switched to custom variants. Doing so, performance of all ships improved except Centurion, so for that ship I kept results obtained with the vanilla variant and thus did all the following series this way: custom variants for all except vanilla “Assault” Centurion (1.2 to 4.2).
Also worth noting, I changed some of the “rules” at one point. From 1.1 to 2.2, the win condition was perfect win, that is without any retreat (allied/enemy) and without any allied ship disabled. Time to win was only measured when this condition was met. Then, from 3.1 to 4.2, I relaxed the condition to what the game considered a victory, and measured time to win in all those situations.
In any cases, further below in the time distribution charts, I used either a continuous line when win condition was met for all 24 runs, or a dashed line otherwise.
Tested shipsSpoiler
vanilla Lasher, “Strike” variant: 8 capacitors, 10 vents, armoredweapons, targetingunit, blast_doors, 2 lightag, 2 reaper, 2 lightmg, 1 lightdualac
custom Lasher: 9 capacitors, 10 vents, fluxcoil, fluxdistributor, reinforcedhull, 2 lightag, 2 sabot_single, 2 lightmg, 1 lightdualac
vanilla Centurion, “Assault” variant: 9 capacitors, 10 vents, blast_doors, 1 annihilator, 2 lightac, 1 lightag, 1 ioncannon, 2 vulcan
custom Centurion: 7 capacitors, 10 vents, reinforcedhull, fluxcoil, fluxdistributor, 1 sabot_single, 2 lightac, 2 lightag, 2 lightmg
vanilla Vigilance, “Standard” variant: 0 capacitors, 8 vents, fluxdistributor, missleracks, eccm, 1 pulselaser, 1 harpoonpod
custom Vigilance: 10 capacitors, 10 vents, reinforcedhull, 1 pulselaser, 1 sabotpod
vanilla Wolf, “Assault” variant: 7 capacitors, 10 vents, fluxdistributor, 1 heavyblaster, 2 harpoon, 1 ioncannon, 2 pdlaser
custom Wolf: 10 capacitors, 10 vents, reinforcedhull, fluxcoil, fluxdistributor, 1 heavyblaster, 2 sabot_single, 2 pdlaser
vanilla Brawler, “Starting” variant: 7 capacitors, 10 vents, fluxbreakers, 2 heavymortar, 2 railgun
custom Brawler: 8 capacitors, 10 vents, reinforcedhull, fluxdistributor, 1 heavymortar, 1 arbalest, 2 sabot
vanilla Tempest, “Attack” variant: 6 capacitors, 10 vents, targetingunit, fluxdistributor, stabilizedshieldemitter, 1 pulselaser, 1 gravitonbeam, 1 harpoon
custom Tempest: 10 capacitors, 10 vents, reinforcedhull, fluxdistributor, 1 heavyblaster, 1 sabot, 1 pdlaser
vanilla Scarab, “Starting” variant: 9 capacitors, 10 vents, blast_doors, fluxdistributor, 2 atropos_single, 1 amblaster, 1 ioncannon, 1 irpulse, 2 pdlaser
custom Scarab: 8 capacitors, 10 vents, reinforcedhull, fluxdistributor, fluxcoil, 2 sabot_single, 1 amblaster, 2 irpulse, 2 pdlaser
vanilla Shrike, “Attack” variant: 9 capacitors, 20 vents, unstable_injector, 1 heavyblaster, 1 sabotpod, 2 irpluse, 1 ioncannon, 2 pdlaser
custom Shrike: 15 capacitors, 20 vents, reinforcedhull, 1 heavyblaster, 1 sabotpod, 2 irpulse, 2 pdlaser
vanilla Medusa, “Attack” variant: 0 capacitors, 19 vents, hardenedshieldemitter, fluxbreakers, 2 railgun, 2 heavyblaster, 2 irpulse, 2 pdlaser
custom Medusa: 14 capacitors, 20 vents, reinforcedhull, fluxdistributor, 2 railgun, 1 heavyblaster, 2 pdlaser, 1 pdburst
vanilla Hyperion, “Attack” variant: 5 capacitors, 10 vents, fluxdistributor, hardenedshieldemitter, hardened_subsystems, 1 pulselaser, 1 phasebeam, 1 heavyneedler
custom Hyperion: 10 capacitors, 10 vents, extendedshieldemitter, unstable_injector, reinforcedhull, 1 ionpulser, 1 miningblaster, 1 heavymg
Tested ships were either used as single ship fleets (1.1 to 2.2), or used as part of 16 DP (3.1 and 3.2) and 25 DP fleets (4.1 and 4.2).
Enemy shipsSpoiler
vanilla Kite (A), “Starting” variant: 9 capacitors, 10 vents, 1 heatseeker, 1 swarmer, 1 lightdualac
vanilla Cerberus, “Starting” variant: 9 capacitors, 10 vents, blast_doors, reinforcedhull, 1 lightac, 1 lightag, 1 vulcan, 1 flak
vanilla Lasher, “Strike” variant: 8 capacitors, 10 vents, armoredweapons, targetingunit, blast_doors, 2 lightag, 2 reaper, 2 lightmg, 1 lightdualac
vanilla Condor, “Support” variant: 8 capacitors, 15 vents, 1 salamanderpod, 2 lightac, 2 talon_wing
vanilla Enforcer, “Assault” variant: 19 capacitors, 20 vents, targetingunit, fluxdistributor, 2 sabot, 2 reaper, 2 heavyac, 1 heavymortar, 2 flak
vanilla Falcon, “Attack” variant: 10 capacitors, 20 vents, autorepair, heavyarmor, 2 heavymg, 2 heatseeker, 2 pulselaser, 4 pdlaser
vanilla Dominator, “Support” variant: 21 capacitors, 30 vents, dedicated_targeting_core, fluxdistributor, 2 mark9, 3 pilum, 3 lightag, 6 vulcan, 2 flak
Enemy ships were used in various fleet compositions, detailed below.
Test series with performance ranking and data visualisation(1.1) Single ship test fleet (vanilla variants) vs
1 Kite (A): Scarab > Tempest > Shrike > Hyperion > Centurion=Lasher > the rest
(1.2) Single ship test fleet vs
1 Kite (A): Tempest=Scarab > Lasher=Shrike=Hyperion > Centurion > the rest
(1.3) Single ship test fleet vs
2 Kite (A): Tempest > Shrike > Scarab > Medusa=Hyperion > the rest
(2.1) Single ship test fleet vs
1 Cerberus: Tempest > Shrike > Scarab > Medusa > Hyperion=Lasher > the rest
(2.2) Single ship test fleet vs
2 Cerberus: Tempest > Shrike > Scarab > Medusa > Hyperion > the rest
(3.1) 16 DP test fleet vs
1 Kite (A) + 1 Lasher + 1 Condor: Tempest > Scarab > the rest
(3.2) 16 DP test fleet vs
1 Falcon: Tempest > Centurion > Scarab > Wolf > the rest
(4.1) 25 DP test fleet vs
1 Kite (A) + 1 Lasher + 1 Condor + 1 Enforcer: Tempest > Scarab > the rest
(4.2) 25 DP test fleet vs
1 Dominator: Tempest > Scarab > Shrike=Lasher=Wolf > Centurion > the rest
CommentsI’d like to remind you: basically only one loadout per ship was tested.
Lasher did OK when in 1v1 situations against weaker ships (Kite and Cerberus) and when outnumbering a strong slow ship (6v1 against Dominator). Lasher was weak when outnumbered.
Centurion did OK in similar situations as Lasher, and had good performance doing 4v1 against Falcon.
Vigilance was not meant to perform well in such tests, so no surprise about its bad performance. Still… maybe it should perform a bit better?
Wolf only performed well when outnumbering opponent, 3v1 against Falcon, and 5v1 against Dominator. A Pulse Laser would probably have increased performance against fighters, frigates and Condor, while lowering performance against cruisers (and Enforcer).
Brawler was at best an OK performer in a few tests, and otherwise pretty weak. Admittedly I don’t really know the ship, also it is entirely possible it wasn’t meant to perform well in such tests. Maybe long range ballistic guns with Harpoons would have worked better?
Tempest was of course the star of the show, no surprise. Best tested ship, results speak for themselves.
Scarab performed really well in all tests. Overall second position, behind Tempest.
Shrike performed well against Kite, Cerberus and Dominator. Shrike was weak in all 16 DP fleet tests and in the first 25 DP fleet test. It should be noted, one Shrike had more Sabots than all single ship fleets except Vigilance which had the same amount.
Medusa performance was disappointing as single ship and in 25 DP fleet tests. It was not part of the 16 DP fleet tests. I do not have a lot of experience with AI-controlled Medusa, so I'm not sure how to interpret those results. I guess a few factors should be considered: it was one of slowest tested ship, it had no missile, and because of its high DP there was less ships on the field in the 25 DP fleet tests.
Hyperion performance was also disappointing. Admittedly I have zero experience with AI-controlled Hyperion, and I gave it my last player-controlled loadout, which may be a cause for the bad performance. That said, such a ship is probably meant to be used either by player or by skilled officer, together with built-in hullmods. With a player-controlled Hyperion using the same loadout I am confident I would win 100% encounters against a lone Falcon, but in the related test above AI Hyperion lost 100% battles. Also, as with Medusa: it had no missile, and because of its high DP there was less ships on the field in the 16 DP and 25 DP fleet tests. So all in all I'm not sure whether those results are relevant or not... except I guess we can see here AI Hyperion may have clear downsides/weaknesses, and that cannot be said about AI Tempest or AI Scarab.
